Background/Objectives: Tumor budding (TB)—clusters of one to five tumor cells at the invasive front—has emerged as a prognostic marker in various cancers. Its prognostic value in head and neck squamous cell carcinoma (HNSCC) is unclear. Methods: We retrospectively analyzed 98 HNSCC patients. The tumor buds were counted on hematoxylin–eosin-stained sections as per the 2016 International Tumor Budding Consensus Conference (ITBCC) guidelines. An optimal cutoff was determined by ROC analysis using excisional lymph nodes and five-year overall survival (OS) as the endpoint, stratifying patients into low- (≤4 buds) and high-risk (>4 buds) groups. The associations with clinicopathological features, OS, and disease-free survival (DFS) were assessed using Kaplan–Meier curves and Cox regression. Results: Among the 98 patients (median follow-up 58 months, range 18–108), 32 (32.7%) died. The optimal TB cutoff was 4.5 (AUC 0.85, 95% CI 0.76–0.93). High TB was associated with poorer five-year OS (26.4% vs. 85.3%). Multivariate Cox regression identified TB and extranodal extension as independent predictors of OS (TB HR: 3.4, 95% CI 1.3–9.2, p = 0.013). In the laryngeal cancer subgroup, TB was associated with worse survival in the univariate analysis (HR 7.5, 95% CI 1.6–35.6, p = 0.011), though this was not significant in the multivariate modeling. High TB independently predicted neck lymph node metastasis (multivariate OR 4.9, 95% CI 1.2–20.5, p = 0.029), which was present in 65.8% of the high-TB vs. 31.7% of the low-TB patients. High TB correlated with advanced AJCC stage and lymphovascular invasion. No clinicopathological factors, including TB, independently predicted DFS, in either the full cohort or the laryngeal subgroup. Conclusions: High tumor budding denotes an aggressive HNSCC phenotype and may guide decisions on elective neck dissection. Its assessment is simple, cost-effective, and potentially valuable for routine pathology, pending external validation.

Background and Objective: This study aims to utilize deep learning methods for the automatic segmentation of cervical lymph nodes in magnetic resonance images (MRIs), enhancing the speed and accuracy of diagnosing pathological masses in the neck and improving patient treatment processes. Materials and Methods: This study included 1346 MRI slices from 64 patients undergoing cervical lymph node dissection, biopsy, and preoperative contrast-enhanced neck MRI. A preprocessing model was used to crop and highlight lymph nodes, along with a method for automatic re-cropping. Two datasets were created from the cropped images—one with augmentation and one without—divided into 90% training and 10% validation sets. After preprocessing, the ResNet-50 images in the DeepLabv3+ encoder block were automatically segmented. Results: According to the results of the validation set, the mean IoU values for the DWI, T2, T1, T1+C, and ADC sequences in the dataset without augmentation created for cervical lymph node segmentation were 0.89, 0.88, 0.81, 0.85, and 0.80, respectively. In the augmented dataset, the average IoU values for all sequences were 0.91, 0.89, 0.85, 0.88, and 0.84. The DWI sequence showed the highest performance in the datasets with and without augmentation. Conclusions: Our preprocessing-based deep learning architectures successfully segmented cervical lymph nodes with high accuracy. This study is the first to explore automatic segmentation of the cervical lymph nodes using comprehensive neck MRI sequences. The proposed model can streamline the detection process, reducing the need for radiology expertise. Additionally, it offers a promising alternative to manual segmentation in radiotherapy, potentially enhancing treatment effectiveness.